UPF Litigation Update: Plaintiffs’ Causation Challenge Comes into Sharper Focus

By Anand Agneshwar and Jocelyn Wiesner

In our prior article, “The Rise of Big Food Litigation,” we described the August 2025 dismissal of Martinez v. Kraft Heinz Co. as “a pause, not an end” to the wave of ultra-processed food (UPF) litigation.[1] That prediction has now been tested against the plaintiff’s best attempt to cure his pleading defects—and the pause has held and the end is perhaps closer. On June 30, 2026, Judge Perez denied Martinez’s motion for leave to amend, holding that his proposed amended complaint remained fatally deficient on causation.

A Bigger Complaint, the Same Hole

After the original complaint was dismissed in August 2025 (“Martinez I”), Martinez moved under Rules 15 and 59(e) to reopen the judgment and file a proposed First Amended Complaint (“PAC”) that tripled the length of the original pleading—to 437 pages—and, for the first time, identified 179 specific products the plaintiff allegedly consumed between 2009 and 2021, along with (in some instances) rough frequency estimates. Despite the added detail, the PAC contained “very few distinctions” between each of the products consumed, generally alleging that “his consumption of a class of products is the cause of his illness.”[2]

Defendants opposed, arguing both that (1) Martinez had known of the facts underlying the PAC—the specific products he consumed—for years, and that his eleventh-hour amendment reflected impermissible “wait-and-see” gamesmanship and (2) that his proposed amendment was futile. Although the court declined to find that Martinez had unduly delayed, it agreed that the proposed amended complaint was futile.

In doing so, the court reaffirmed its central holdings from Martinez I. First, the court held that despite the added length, the PAC “fails to establish that each of the 179 products named necessarily contributed to [the plaintiff’s] injuries.”[3] Applying the but-for test for causation, the court asked whether, if one or more defendants or products was removed from the equation, the outcome would change. The court concluded that it wouldn’t and therefore that the complaint did not adequately plead specific causation. The PAC’s allegations of “biological plausibility” and ingredient-by-ingredient “association” with disease risk, the court explained, showed only correlation, and “correlation is not causation.” The court likewise rejected plaintiff’s reliance on Gray v. Abbott Laboratories, an unpublished opinion in the infant formula litigation. As the court explained, there is a material difference between a single contaminated, recalled product consumed immediately before acute illness and 179 products from eleven different manufacturers allegedly consumed over more than a decade before a chronic diagnosis.

Second, the court rejected the plaintiff’s attempt to invoke alternative liability and market-share liability as a way around the causation problem. Both doctrines, the court held, require that the defendants’ products be fungible or otherwise identical in their harmful characteristics—a requirement flatly inconsistent with a complaint that itself catalogs dozens of different allegedly harmful ingredients across different products with different consumption frequencies. As the court put it, “nowhere does the [PAC] suggest that each of the 179 named products are equally as harmful as the others,” and Pennsylvania courts have shown no willingness to extend these doctrines beyond truly identical products.[4] Moreover, “[a]lternative liability does not apply where actual cause has or cannot be established.”[5]

Why This Matters: Causation is Not Merely a Pleading Issue

The ruling is not merely ruling on a pleading technicality. UPF litigation to date rests on epidemiological associations—population-level correlations between processed-food consumption and chronic disease—that were not designed to answer the question a tort case must answer: “Did this product cause this plaintiff’s disease?”, or even, “Is this product capable of causing this disease?” Converting such associational data into individualized legal causation—before the underlying science can reliably attribute a specific disease to a specific product or ingredient—would let litigation outrun epidemiology. The court’s insistence on but-for causation, and its refusal to let alternative or market-share liability substitute for it, keeps the causation inquiry tethered to what the science can actually support.

A plaintiff who names the specific brands he consumed and alleges (as later complaints have) approximate frequencies and start dates still faces the same structural problem: a chronic, multifactorial disease that the complaints attribute to years of consumption of dozens of different products from multiple manufacturers. This “creates a unique challenge for plaintiffs” which cannot be overcome through mere pleading changes. This reasoning should provide strong guidance in other personal injury cases.

While the analysis transfers less directly to government enforcement actions, it is not irrelevant. As discussed in our prior article, consumer protection and public nuisance theories still require general causation—even if not specific causation. The Martinez II court’s broader skepticism of treating “UPF” as an undifferentiated category with a single causal mechanism has persuasive force where a government plaintiff seeks to attribute discrete public health costs to the industry as a whole, or to any one defendant’s market share, without a more particularized causal or apportionment theory. To be sure, Martinez II noted that the PAC “may plausibly allege that UPFs, or certain ingredients within them, are generally capable of increasing the risk” of certain diseases.[6] But it was also careful to point out that “correlation between the rise in UPFs” generally and the rise in certain pediatric illnesses “does not amount to causation” attributable to any particular product. So, while its focus was on specific—not general—causation, it previews some of the hurdles government actors will face. Defense counsel litigating government cases should watch for opportunities to import that reasoning.

The Docket Keeps Growing, But Will Soon Be Tested

The wave of copycat personal injury suits continues even as the pleading bar holds firm. That said, we may soon be able to see if other courts will adopt the Martinez reasoning. In Lawton v. Kraft Heinz Co., for example, defendants filed a notice of supplemental authority citing Martinez in support of their pending motion to dismiss.[7] And, in Muthusami v. Kraft Heinz Co., the plaintiff filed a First Amended Complaint one week after Martinez II.[8] Notably, that amended complaint narrows the alleged injury to a single disease and identifies a smaller universe of products (roughly 80, compared to Martinez’s 179). At the same time, however, the amended complaint’s core structure remains essentially unchanged from the pleading style Martinez rejected. Each cause of action is asserted collectively against every defendant with nearly identical boilerplate paragraphs repeated for each named defendant grouping. In other words, the amended complaint leaves intact the same multi-product, multi-defendant structure that doomed the Martinez complaint. Finally, Martinez himself has filed a notice of appeal, giving the Third Circuit an opportunity to weigh in.

Conclusion

Martinez II confirms what the original dismissal foreshadowed: specific causation, not the sheer volume of product-specific detail, is the central and likely insurmountable obstacle for personal injury plaintiffs pursuing an aggregated, multi-defendant UPF theory. While government actions remain more challenging for defendants, they too will face causation challenges. Meanwhile, the regulatory question that will ultimately define what counts as a “UPF” remains open at the federal level, even as states continue to fill the gap.[9] Defense counsel should continue to track how these threads—causation, pleading specificity, and regulatory definition—develop in tandem across the UPF docket.

[1] Anand Agneshwar & Jocelyn Wiesner, “The Rise of Big Food Litigation: Is a Landmark Decision in Pennsylvania Only the Beginning?” in Top Food and Drug Cases, 2025, Food & Drug Law Institute (June 24, 2026), https://www.fdli.org/2026/06/the-rise-of-big-food-litigation/.

[2] Martinez v. Kraft Heinz Co., Inc., No. 2:25-cv-00377 (E.D. Pa. June 30, 2026), ECF No. 162 (“Martinez II”) at 4, 15.

[3] Martinez II at 2.

[4] Id. at 16.

[5] Id. at 15.

[6] Id. at 12.

[7] See Lawton v. Kraft Heinz Co., No. 1:26-cv-44 (S.D. Miss. July 14, 2026), ECF No. 117. Notably, the magistrate judge in Lawton granted defendants’ motion to stay discovery pending resolution of the omnibus motion to dismiss. Id. at ECF No. 111. Because the motion to dismiss raises a personal jurisdiction challenge as to two defendants (Mondelez and Post Holdings), the district’s local rules triggered an automatic stay as to those defendants; the court extended that stay to all defendants “in the interest of judicial economy,” while reserving the right to permit limited jurisdictional discovery.

[8] Muthusami v. Kraft Heinz Co., No. 6:26-cv-00113 (M.D. Fl. July 7, 2026), ECF No. 110.

[9] FDA and USDA’s effort to finalize a uniform federal definition of “ultra-processed food”—first launched via a July 2025 Request for Information as discussed in our prior article—remains outstanding; as of mid-June 2026, HHS Secretary Robert F. Kennedy Jr. said only that “[w]e anticipate that we’re going to be able to do a final publication over the next couple of months.”

 

FDLI is a nonprofit membership organization that offers education, training, publications, and professional networking opportunities in the field of food and drug law. As a neutral convener, FDLI provides a venue for stakeholders to inform innovative public policy, law, and regulation. Articles and any other material published in Update represent the opinions of the author(s) and should not be construed to reflect the opinions of FDLI, its staff, or its members. The factual accuracy of all statements in the articles and other materials is the sole responsibility of the authors.